Using a combination of single subject and group designs the implications of blood glucose discrimination training on diabetic control in adult insulin dependent diabetics will be assessed. Study 1 will evaluate the effect of providing systematic feedback regarding blood glucose levels on diabetic individuals' ability to discriminate changes in their blood glucose concentrations. Blood glucose levels will be measured using home blood glucose monitors. In a reversal design, two groups of subjects will be asked to estimate their blood glucose levels at three daily assessments. During the treatment condition they will be provided with feedback regarding the accuracy of their estimation. Additionally, following the second exposure to feedback conditions, one group of subjects will be given training designed to enhance maintenance of blood glucose discrimination. This study will focus on the accuracy with which diabetics can learn to discriminate their blood glucose levels, as well as, whether this behavior change will be continued following maintenance training. Study 2 will evaluate the impact of blood glucose discrimination on long term diabetic control. Using a procedure similar to that employed in study 1, subjects will be trained to discriminate changes in blood glucose levels. Overall diabetes control will be assessed using hemoglobin A1c measure of metabolic control. It is hoped that these investigations will determine whether diabetic individuals can learn to accurately detect changes in their blood glucose levels, whether these skills can be maintained, and if discriminating blood glucose levels results in an improvement in long term diabetic control.